Data Tracking Systems - Biomedical Informatics in Translational Research - page 120

7.3.1 General Requirements
A major general requirement is the extensibility of the system. Extensibility refers to
the ability to add new functions or modify the existing ones without having to make
major changes to the infrastructure itself [53]. As a general requirement for IT prod-
ucts, extensibility is particularly important for data tracking systems for biomedical
informatics research. The dynamic nature of the field often demands modifications
of existing protocols and the addition of new experimental platforms and thus new
protocols. The data tracking system needs to be modified accordingly without
changing its backbone structure.
The data tracking system can be workflow based, because the workflow concept
is relatively intuitive to end users. It should track the major operational steps, and
capture the data generated by the instrument(s). The QA/QC results should be
tracked as well.
An easily configurable system is especially desirable. System administrators
should be able to assign role-based privileges to all users [54], for example, to allow
the principal investigator to design the experiment, allow project managers to select
the protocol and specimens to start the experimental workflow, and allow labora-
tory research associates to carry out the experiment following the workflow. A data
tracking system should always store the data with a time stamp and user stamp, and
such stamps can be very useful for troubleshooting.
Users with high privileges, such as the system manager or lab manager, should
be allowed to develop a new workflow or revise an existing one by inserting a new
step or deleting or modifying an existing step. In doing so, all the changes need to be
audited for future reference. Meanwhile, the system should store the old data and
the new data in such a way that, after necessary modifications, the reporting utility
will be able to report both the old data and the new data in the same way, for exam-
ple, by allowing nulls to be reported if a step is inserted or deleted. In addition, if the
data in the tracking system are to be loaded into a data storage system such as a data
warehouse, then, after proper modifications, the data exporting utility should be
able to handle the old data and the new data the same way.
7.3.2 Front-End Requirements
Because end users of the data tracking system may have limited computer skills, it is
important for the interfaces to be user friendly and intuitive. Most likely this is the
only source of information the end users have to judge the quality of the system. If
the end users do not like the system, chances are, it will not make it through the
deployment stage.
Flexibility is also important in the user interface. This does not mean that all
users should be able to modify the system, but privileged users should be allowed to
do so as discussed earlier. The front end should also allow problem reporting and
solving with a type of two-way mechanism instead of just a one-way reporting sys-
tem as is seen in many LIMS where, when a problem occurs the workflow leads to a
problem reporting step and ends there. In such cases, even when a solution to the
problem is later found, the previously reported step cannot be corrected.
Barcode reading is necessary to keep track of the questionnaires, specimens,
sample vials, plates, and so forth. Barcode reading steps should be properly planned